Influence of the Real Background Signal on the Linearity of the Stern-Volmer Calibration for the Determination of Molecular Oxygen with Optical Sensors

The role of light emission components unrelated to the Stern-Volmer (SV) kinetic model, in the curvature of the Stern-Volmer model, was evidenced and the discrepancy often encountered between calibrations with light intensity, I, and luminescence lifetime,τ, was explained by means of mathematical models applied to experimental data. Sensing membranes prepared embedding tris-4,7-diphenyl-1,10-phenanthroline (Ru(dpp)), platinum(II) meso-tetraphenylporphyrin (PtTPP), and palladium(II) meso-tetra(pentafluorophenyl)porphyrin (PdFTPP) in a polysulfone matrix were used to confirm the proposed theory. Multisite hypotheses or other correction procedures must be invoked only after a careful evaluation of the background light really present in the system. A new form of the SV equation having the advantage to obtain the K′SV and the effective background with a simple unweighted regression method was proposed and compared to the “classical” SV requiring the knowledge of the background and a weighted regression. Experimental results confirmed the theory for what concerns the concentration limit for the determination of oxygen which may be theoretically determined. Ru(dpp)OS, PtTPP, and PdFTPP membranes may determine oxygen mixtures up to 100, 60.7, and 19.0%, respectively, with a resolution of 2%. Narrower working intervals are explorable if better resolution is required.